A novel controlled release formulation for the anticancer drug …drugsanddisease.synthasite.com/resources/novel-controlled... · 2010-07-23 · A novel controlled release formulation

Journal of Controlled Release 86 (2003) 33–48www.elsevier.com/ locate/ jconrel

A novel controlled release formulation for the anticancer drugpaclitaxel (Taxol ): PLGA nanoparticles containing vitamin E

TPGSa a,b ,*L. Mu , S.S. Feng

aDivision of Bioengineering, The National University of Singapore, 9 Engineering Drive 1, Singapore 117576,SingaporebDepartment of Chemical and Environmental Engineering, The National University of Singapore, 9 Engineering Drive 1,

Singapore 117576,Singapore

Received 26 April 2002; accepted 20 September 2002

Abstract

Paclitaxel (Taxol ) is one of the best antineoplastic drugs found from nature in the past decades. Like many otheranticancer drugs, there are difficulties in its clinical administration due to its poor solubility. Therefore an adjuvant calledCremophor EL has to be employed, but this has been found to cause serious side-effects. However, nanoparticles ofbiodegradable polymers can provide an ideal solution to the adjuvant problem and realise a controlled and targeted deliveryof the drug with better efficacy and fewer side-effects. The present research proposes a novel formulation for fabrication ofnanoparticles of biodegradable polymers containing d-a-tocopheryl polyethylene glycol 1000 succinate (vitamin E TPGS orTPGS) to replace the current method of clinical administration and, with further modification, to provide an innovativesolution for oral chemotherapy. In the modified solvent extraction/evaporation technique employed in this research, theemulsifier /stabiliser /additive and the matrix material can play a key role in determining the morphological, physicochemicaland pharmaceutical properties of the produced nanoparticles. We found that vitamin E TPGS could be a novel surfactant aswell as a matrix material when blended with other biodegradable polymers. The nanoparticles composed of variousformulations and manufactured under various conditions were characterised by laser light scattering (LLS) for size and sizedistribution, scanning electron microscopy (SEM) and atomic force microscopy (AFM) for morphological properties, X-rayphotoelectron spectroscopy (XPS) for surface chemistry and differential scanning calorimetry (DSC) for thermogramproperties. The drug encapsulation efficiency (EE) and the drug release kinetics under in vitro conditions were measured byhigh performance liquid chromatography (HPLC). It was concluded that vitamin E TPGS has great advantages for themanufacture of polymeric nanoparticles for controlled release of paclitaxel and other anti-cancer drugs. Nanoparticles ofnanometer size with narrow distribution can be obtained. A drug encapsulation efficiency as high as 100% can be achievedand the release kinetics can be controlled. 2002 Elsevier Science B.V. All rights reserved.

Abbreviations: AFM, atomic force microscopy; DCM, dichloromethane; DSC, differential scanning calorimetry; FDA, US Food and DrugAdministration; HPLC, high performance liquid chromatography; LLS, laser light scattering; PBS, phosphate-buffered saline; PLGA, poly(lactic-co-glycolic acid); PVA, polyvinyl alcohol; SEM, scanning electron microscopy; Vitamin E TPGS or TPGS, d-a-tocopherylpolyethylene glycol 1000 succinate; XPS, X-ray photoelectron spectroscopy

*Corresponding author. Present address: Department of Chemical and Environmental Engineering, The National University of Singapore,9 Engineering Drive 1, Singapore 117576, Singapore.

E-mail address: [email protected](S.S. Feng).

0168-3659/02/$ – see front matter 2002 Elsevier Science B.V. All rights reserved.PI I : S0168-3659( 02 )00320-6

mailto:[email protected]

34 L. Mu, S.S. Feng / Journal of Controlled Release 86 (2003) 33–48

Keywords: Biodegradable polymer; Drug delivery; d-a-Tocopheryl polyethylene glycol 1000 succinate; Emulsifier; Surfactant stabiliser;Taxol

1 . Introduction porosity, and the fabrication conditions [14,15]. Thefabrication technique for micro/nanoparticles should

The application of biodegradable polymeric par- be chosen based on the nature of the polymer, theticles in the scale of micrometers and nanometers as drug, the intended use, and the duration of thea controlled release dosage form of anticancer drugs therapy [2,16,17]. The resulting pharmaceuticalhas generated immense interest. Due to their rela- properties of the micro/nanoparticles may be de-tively large size, however, the microspheres were not termined by various factors which include the nature,appropriate to direct the drug to target tissues or cells solubility and loading of the drug; the polymer type,via systemic circulation or across the mucous mem- composition, and molecular weight; the property ofbrane [1–3]. Nanoparticles, instead, were successful- organic solvent; the concentration and mixed ratio ofly used for systemic, oral, pulmonary, transdermal the water and the oil phases; the nature and theand other administration routes for various purposes concentration of the emulsifier; the mechanicalincluding drug targeting, enhancement of drug bio- strength of stirring/agitation; and other conditionsavailability and protection of drug bioactivity and such as temperature and pH, etc. Among them, thestability [4–6]. Also, nanoparticles can improve the matrix material and the emulsifier /stabiliser /additivebioavailability of poorly absorbed drugs, thus en- can play a key role.hancing oral delivery [3,4]. Moreover, the nanoparti- The present research proposes d-a-tocopherylcles are able to permeate cells for cellular internalisa- polyethylene glycol 1000 succinate (vitamin E TPGStion and connective tissue permeation and so deliver or TPGS) as a novel emulsifier as well as a novelthe drug efficiently to the targeted tissue without matrix material when blended with other biodegra-clogging capillaries [7,8]. The ability of nanoparti- dable polymers in the preparation of nanoparticlescles to improve drug diffusion through biological for controlled release of anticancer drugs with pa-barriers is a typical benefit for the delivery of clitaxel as a prototype, which is one of the bestanticancer agents. The enhanced endocytic activity natural antineoplastic drugs of the past decades. Theand leaky vasculature in the tumor could result in therapeutic efficacy of this drug has been limited dueaccumulation of intravenously administered to its poor aqueous solubility. In current clinicalnanoparticles [9]. Some studies have indicated that administration of paclitaxel, an adjuvant called Cre-nanoparticle-bound antitumour agents showed pro- mophor EL is needed, which has been proved to belonged drug retention in tumours, reduction in responsible for most serious side-effects of thetumour growth and prolonged survival of tumour- dosage form such as hypersensitivity reaction, ne-bearing animals [10–13]. phrotoxicity, neurotoxicity and cardiotoxicity [18–

The polymer matrix of the nanoparticles must 24]. The nanoparticles can provide a solution for themeet several requirements such as biocompatibility, problems caused by Cremophor EL and with furtherbiodegradability, mechanical strength, and ease of modification, promote oral chemotherapy due to theirprocessing. The best known class of biodegradable extremely small size and adhesive properties.materials for controlled release are the poly (lactide- A widely used emulsifier in the preparation ofco-glycolide)s (PLGAs). In addition, a wide variety micro/nanoparticles is poly (vinyl alcohol) (PVA)of biopolymers such as bovine serum albumin [25–29]. To our knowledge, there have been no(BSA), human serum albumin (HSA), collagen, comprehensive studies on applying vitamin E TPGSgelatine, and hemoglobin have been studied for their as an emulsifier, let alone using it as matrix material.application to drug delivery systems [2]. The only exception is a recent publication by the

Various drug release profiles can be achieved by authors, in which they initiated and evaluated thecontrolling the molecular weight, the copolymer feasibility and advantages of TPGS utilised as emul-ratio, the drug loading, the microparticle size and sifier in the solvent evaporation/extraction technique

L. Mu, S.S. Feng / Journal of Controlled Release 86 (2003) 33–48 35

[25]. The present work more intensively investigated used throughout the experiment. The in vitro releaseTPGS used as emulsifier and further pursued the measurement was carried out at pH 7.4 and 378C inpossibility of applying TPGS as a matrix material phosphate-buffered saline (PBS), which was pur-blended with PLGA for the manufacture of chased from Sigma. All other chemicals used werenanoparticles for clinical administration of paclitaxel. of reagent grade.The preparation, characterisation and in vitro releasekinetics of the nanoparticles of this novel formula- 2 .2. Nanoparticle preparationtion were studied. The different formulations withvarious ratios of the oil phase, the aqueous phase, the The paclitaxel loaded nanoparticles were fabri-polymer material and the TPGS either as emulsifier cated by a modified oil-in-water single emulsionor as a component of the matrix material were solvent evaporation/extraction technique [25].evaluated and optimised. The results demonstrated Briefly, known amounts of mass of polymer andthat vitamin E TPGS could be an excellent emulsifier paclitaxel were added into DCM, which was suitablyfor fabrication of polymeric nanoparticles, achieving stirred to ensure that all material was dissolved. Thevery good emulsifying effects and high drug en- solution of organic phase was slowly poured into thecapsulation efficiency. The desired size and size stirred aqueous solution with or without emulsifierdistribution, surface morphology, and in vitro release and sonicated simultaneously at 50 W in pulse modekinetics can be obtained. Moreover, this is the first (Misonix, USA). The formed o/w emulsion wastime in the literature that TPGS has been used as a gently stirred at room temperature (228C) by amatrix material for nanoparticle preparation, having magnetic stirrer overnight to evaporate the organica self-emulsifying effect in the emulsification pro- solvent. The resulting sample was collected bycess. centrifugation (10,000 rpm, 10 min, 168C; Eppen-

dorf model 5810R, Eppendorf, Hamburg, Germany)and washed once or twice with deionised water for

2 . Materials and methods some samples. The produced suspension was freeze-dried (Alpha-2, Martin Christ Freeze Dryers, Ger-

2 .1. Materials many) to obtain a fine powder of nanoparticles,which was placed and kept in a vacuum dessicator.

Poly(DL-lactide-co-glycolide) (PLGA; L/G550/ The loading ratio of paclitaxel for the preparation50, MW 40,000–75,000; L/G575/25, MW 90,000– was 1–10%.120,000; and L/G585/15, MW 90,000–120,000),poly(DL-lactide) (PLA; MW 106,000), polyvinyl 2 .3. Encapsulation efficiencyalcohol (PVA; MW 30,000–70,000; the viscosity ofa 4% solution was 4–6 cp at 208C; the degree of The drug entrapped in the nanoparticles washydrolysis was 87–90%) were purchased from determined in triplicate by HPLC (Agilent LC1100,Sigma (USA). Paclitaxel of 99.8% purity was pur- Agilent Technologies, Singapore). A reverse phase

chased from Yunnan Hande Biotechnology, China. Inertsil ODS-3 column (15034.6 mm i.d., pore sizeVitamin E TPGS was a gift from Eastman Chemical, 5mm, GL Science, Tokyo, Japan) was used. A 3-mgUSA. Methylene chloride (dichloromethane, DCM, sample of nanoparticle powder was dissolved in 1 mlanalytical grade) was purchased from Mallinckrodt of DCM and 5 ml of acetonitrile–water (50:50) was(Mallinckrodt Laboratory Chemicals, Mallinckrodt then added. A nitrogen stream was introduced toBaker, USA). Acetonitrile used as mobile phase in evaporate the DCM until a clear solution washigh performance liquid chromatography (HPLC) obtained. The solution was filtered into a vial forwas purchased from EM Science (ChromAR, HPLC HPLC detection of the paclitaxel concentration. Thegrade, Mallinckrodt Baker, USA). Ultra-high pure mobile phase consisted of a mixture of acetonitrilewater produced by UHQ Water Purification System and water (50:50, v /v), and was delivered at a flow(USF-ELGA lab water, Millipore, Singapore) was rate of 1.00 ml /min with a pump (HP 1100 highutilised for HPLC analysis. Deionised water was pressure gradient pump, Agilent Technologies, Sing-


apore). A 50-ml aliquot of the sample was injected nanoparticles. SEM requires coating of the samplewith an autoinjector (HP 1100 Autosampler, Agilent with platinum, which was performed in an Auto FineTechnologies, Singapore). The column effluent was Coater (JFC-1300, JEOL USA). AFM was conducteddetected at 227 nm with a variable wavelength with Nanoscope IIIa in the tapping mode. Thedetector (VWD). The calibration curve for the nanoparticle sample was mounted on metal slabsquantification of paclitaxel was linear over the range using double-sided adhesive tapes and scanned byof standard concentration of paclitaxel at 10–60,000 the AFM maintained in a constant-temperature and

2ng/ml with a correlation coefficient ofR 51.000. vibration-free environment.The solvent for calibration was the mixture ofacetonitrile and water (50:50, v /v). 2 .4.3. DSC analysis

Correction of the calculated encapsulation ef- The thermogram characteristics of selected batchesficiency may be needed in case of inefficient ex- of nanoparticles were determined by differentialtraction [29]. To decide whether correction is scanning calorimetry thermogram analysis (DSC,needed, the recovery efficiency factor of the ex- 2920 Modulated, Universal V2.6D, TA Instruments,traction procedure on encapsulation efficiency was USA) on the glass transition temperatures (T ) org

determined as follows. The same amount of pure melting point (T ). The following steps were takenm

paclitaxel as that loaded in the nanoparticles and during the process. Samples (8 mg) were equili-3.0–5.0 mg of placebo nanoparticles or polymer brated at210 8C and purged with pure dry nitrogenwere dissolved in 1 ml of DCM. Then 5 ml of at a flow rate of 40 ml /min. The nitrogen was heatedacetonitrile–water (50:50) was added. The same to 1208C at 208C/min, after which it was heldprocedure as described above was carried out. The isothermally for 3 min. The samples were cooledresulting factor was 100%, which means that 100% back to210 8C at the same rate. After 5 min ofof the original amount of the paclitaxel could be isothermal stage, the second heating cycle proceededdetected. No correction was needed. The encapsula- at 58C/min temperature ramp speed to 1208C. Thetion efficiency of paclitaxel was obtained as the mass T of polymer was obtained by taking the mid-pointg

ratio between the amount of paclitaxel incorporated of the slope during glass transition. In the presentin nanoparticles and that used in the nanoparticle research, two heating cycles were conducted. Indiumpreparation. was used as the standard reference material to

calibrate the temperature and energy scales of the2 .4. Nanoparticle characterisation DSC instrument. As a control, the pure material was

analysed to observe the change of theT or T .g m

2 .4.1. Size and size distributionThe particle size and size distribution of the 2 .4.4. Surface analysis

prepared nanoparticles were measured by laser light X-ray photoelectron spectroscopy (XPS, Kratosscattering (LLS, 90 Plus Particle Sizer, Brookhaven Axis HSi, Kratos Analytical, Shimadzu, Japan) wasInstruments, USA). The dried powder samples were utilised to analyse the surface chemistry of thesuspended in deionised water and sonicated before nanoparticles. The angle of X-ray was 908. Themeasurement. The obtained homogeneous suspen- analyser was used in fixed transmission mode withsion was examined to determine the volume mean pass energy of 80 eV for the survey spectrumdiameter, size distribution and polydispersity. covering a binding energy range from 0 to 1200 eV.

Peak curve fitting of the C1s (atomic orbital 1s of2 .4.2. Morphology carbon) envelope was performed using the software

Scanning electron microscopy (SEM, JSM-5600 provided by the instrument manufacturer.LV, JEOL USA) and atomic force microscopy (AFM,Multimode� Scanning Probe Microscope, Digital 2 .5. In vitro release studyInstruments, USA) were employed to determine theshape and surface morphology of the produced The release rate of paclitaxel from the nanoparti-


cles was measured in PBS medium (pH 7.4) byHPLC in triplicate. Paclitaxel loaded nanoparticles(10 mg) were suspended in 10 ml of buffer solutionin screw capped tubes and placed in an orbital shakerbath (GFL-1086, Lee Hung Technical, Bukit BatokIndustrial Park A, Singapore), which was maintained

21at 378C and shaken horizontally at 120 min . Aftera particular time interval, the tubes were taken out ofthe water bath and centrifuged at 11000 rpm for 15min. The precipitated nanoparticles were resuspend- Fig. 1. Chemical structure of vitamin E TPGS.ed in 10 ml of fresh buffer and then put back in theshaker bath. The supernatant was taken for analysisof paclitaxel concentration, extracted first with 1 ml water as well as being soluble in oil. The specialof DCM, followed by adding 3 ml of the mixture of structure–property relationship of TPGS suggests itsacetonitrile and water (50:50, v /v), and then evapo- potential use as an emulsifier for various oil–waterrated under a stream of nitrogen until a clear solution immiscible systems. We further realised that, blendedwas obtained. HPLC analysis was conducted as with hydrophobic polymers such as PLA and PLGA,previously described. Similarly to the measurement TPGS could be used as a component of the matrixof encapsulation efficiency, the extraction procedure material to improve the controlled release property ofneeds to be analysed for the extraction recovery nanoparticles. This is a unique characteristic ofefficiency due to inefficient recovery. Known mass TPGS superior to many other emulsifiers such asof pure paclitaxel was dealt with the same procedure PVA.as mentioned above. The determined factor was Our result showed that used as emulsifier added77.5%, which meant that the obtained extraction into the water phase in the microencapsulatingsolution contained 77.5% of the original amount of process, TPGS was very effective at improving thepaclitaxel after the related process. The data obtained emulsification process for microencapsulation. Infor analysis of the in vitro release were corrected fabrication of nanoparticles by the solvent extrac-accordingly. tion/evaporation technique, the concentration needed

for the traditional emulsifier PVA was normally atleast 1% (w/v) [25–28]. However, under the same

3 . Results and discussions fabrication conditions, the required amount of TPGSis only 0.015% (w/v), an amount 67 times less than

3 .1. A novel formulation of nanoparticles for the PVA but with the same emulsifying effects. Also,controlled release of paclitaxel blended with biodegradable polymers such as PLGA,

TPGS could be adopted as a matrix material ofStructurally, vitamin E TPGS has a dual nature, nanoparticles, which was added and dissolved in the

similar to an amphiphile, with part of the molecule oil phase. The mixture of PLGA and TPGS has aexhibiting lipophilicity and another part exhibiting self-emulsifying effect, which can form nanoparticleshydrophilicity, which is necessary for use as a with no need to add another surfactant stabiliser. Insurface-active agent. Although the exact portion of both cases, the resulting nanoparticles have thethe hydrophilic polar head and the lipophilic alkyl desired size with narrow size distribution and thetail may not be elucidated obviously, it is accepted desired release kinetics. The drug encapsulationthat the polyethylene glycol portion behaves as the efficiency could be as high as 100% by optimisingpolar head while the tocopherol succinate portion the formulation. XPS investigation showed that thebehaves as the lipophilic tail (Fig. 1). Moreover, not surface of the fabricated nanoparticles was domi-only does the TPGS molecule possess a bulky shape nated by the TPGS molecules, which demonstratedand large surface area, but it is also miscible with the emulsification role of TPGS. Our work indicated


the most significant advantage by employing vitamin centration of emulsifier. Moreover, samples M1–M3E TPGS either as an efficient stabiliser or as a were produced by applying a blended mixture ofcomponent of the matrix material. This achievement TPGS and PLGA as matrix material with no othersignificantly improves the solvent extraction/evapo- emulsifier or surfactant stabiliser (self-emulsified).ration technique for fabrication of nanoparticles.

3 .2.1. Morphology of nanoparticles3 .2. Formulation optimisation Under SEM and AFM observation (Figs. 2 and 3),

the nanoparticles all had a fine spherical shape withA number of nanoparticle samples were fabricated various degrees of smooth surface. The AFM tech-

using different types of PLGA including PLA, nique was used to study the detailed morphology ofPLGA (85:15), PLGA (75:25) and PLGA (50:50). nanoparticles, as the prepared nanoparticles are tooVitamin E TPGS of various concentrations was small to be closely investigated by SEM due to itsutilised as emulsifier. The nanoparticles with TPGS limited magnification. The AFM images reveal theblended with PLGA as matrix material were also fine structure of the nanoparticle surface as showedfabricated by changing the ratio of polymer and in Fig. 3. They gives clear 3D morphological imagesTPGS. The basic characteristics of the products of spherical nanoparticles of sub-300 nm diameter,involved in the formulation studies are presented in and confirmed that there was no aggregation orTable 1. Samples E1–E4 were prepared by using adhesion among the nanoparticles. Furthermore, theTPGS as emulsifier at high concentration (0.06%, surface morphology of the nanoparticles could bew/v). Samples E5–E8 and E13–E16 were made with seen closely from the AFM images. It was noticeablea medium concentration of TPGS (0.03%, w/v), from the zoom-in picture that caves and/or crackswhile samples E9–E12 employed a low TPGS existed on the particle surfaces and single particlesconcentration (0.015%, w/v). Samples E7, E13, E14 showed a certain roughness on their surface althoughand E8, E15, E16 were prepared by varying the multi-particle images gave relatively smoother sur-polymer concentration and using a medium con- face morphology. Although the topography of the

Table 1The nanoparticles products and their properties

Sample Polymer TPGS Polymer Paclitaxel Mean Poly- Recovery EncapsulationNo. concentration concentration loading diameter dispersity yield (%) efficiency (%)

(%, g/ml) (%, g/ml) (%, w/w) (nm)6S.E.

E1 PLA 0.06 0.125 – 979.06257.7 0.005 38.5 –E2 PLGA (85:15) 0.06 0.125 – 1764.16141.1 0.184 38.9 –E3 PLGA (75:25) 0.06 0.125 – 914.86380.1 0.186 27.1 –E4 PLGA (50:50) 0.06 0.125 – 686.16251.7 0.373 55.9 –E5 PLA 0.03 0.125 2.4 589.06244.5 0.326 44.7 43.0E6 PLGA (85:15) 0.03 0.125 – 1028.26371.8 0.288 55.8 –E7 PLGA (75:25) 0.03 0.125 2.4 272.56169.5 0.245 41.7 50.4E8 PLGA (50:50) 0.03 0.125 2.4 515.56317.9 0.005 64.3 53.2E9 PLA 0.015 0.125 – 846.76348.7 0.243 25.9 –E10 PLGA (85:15) 0.015 0.125 – 567.46362.6 0.277 21.4 –E11 PLGA (75:25) 0.015 0.125 – 699.36286.9 0.005 46.9 –E12 PLGA (50:50) 0.015 0.125 – 895.46318.8 0.052 31.3 –E13 PLGA (75:25) 0.03 0.188 0.83 444.5670.8 0.005 40.5 49.0E14 PLGA (75:25) 0.03 0.25 0.62 654.6680.9 0.005 56.7 80.3E15 PLGA (50:50) 0.03 0.188 0.83 636.4690.6 0.088 47.4 54.3E16 PLGA (50:50) 0.03 0.25 0.62 369.1680.8 0.230 37.7 83.8M1 PLGA (50:50) PLGA/TPGS52/1 9.9 552.3681.4 0.005 46.7 |100M2 PLGA (50:50) PLGA/TPGS51/1 9.9 622.16102.4 0.005 43.6 |100M3 PLGA (50:50) PLGA/TPGS51:2 9.9 844.36115.6 0.005 29.7 |100


(d) Nanoparticles prepared by using TPGS part of matrix material

Fig. 2. SEM images of nanoparticles prepared by different molecular weight of polymer with different concentration of TPGS as emulsifieror by blended mixture of PLGA (50:50) and TPGS without using any other emulsifier.


nanoparticles may be complex, the roughness and thecaves observed on the surface could provide physicalevidence of diffusion release mechanism. The rela-tively smooth surface supported the assumption thatthe release of drug from nanoparticles might becaused by both diffusion and matrix erosion.

3 .2.2. Particle size and size distributionThe mean size averaged by particle volume and

polydispersity of all samples were determined andare listed in Table 1. Fig. 4 illustrates the particlessize distribution. It can be observed that the size ofparticles was in the range 300–1000 nm. It wasshown that, in general, the size of nanoparticles wassmaller when fabricated with 0.03% (w/v) TPGS asemulsifier (samples E5–E8, E13–E16), and thepolydispersity at this emulsifier concentration wasnarrower than those of the nanoparticles fabricated athigher concentration (0.06%, samples E1–E4) andlower concentration (0.015%, samples E9–E12) ofTPGS. This is easy to understand: the emulsifierremains at the interface separating the oil and waterphases. Smaller particles have larger total surfacearea and thus need more emulsifier. Too littleemulsifier would result in no nanoparticles formedand too much emulsifier may result in particleaggregation.

As regards the composition of PLGA, when theL/G ratio decreased, smaller nanoparticles wereobtained. The size distribution became narrower aswell. This result was more notable for PLGA (50:50)(samples E8, E15, E16) and PLGA (75:25) (samplesE7, E13, E14) polymers although other properties ofthe nanoparticles were quite similar. Further, in orderto investigate the effects of polymer concentration onthe formulation properties, different concentrationsof polymer in the oil phase were used to fabricate thenanoparticles. For PLGA (75:25), the nanoparticlesize increased slightly with increasing concentrationof the polymer used, and the polydispersity de-creased to a better value of 0.005. However, thetrend was the opposite for PLGA (50:50). Theparticle size decreased with increasing polymerconcentrations while the polydispersity worsenedslightly. This may be attributed to the differentnatures of the two co-polymers used. As the ratio ofFig. 3. AFM images of nanoparticles prepared with TPGS asL/G is decreased, the hydrophilicity of the polymeremulsifier: (a) multi-particles; (b) single particle; (c) zoom-in of

the nanoparticle surface. increases, and thus the interaction amongst the


Fig. 4. Size distribution of produced nanoparticles under various experiment parameters: (a) TPGS as emulsifier with high concentration(0.06%); (b) TPGS as emulsifier with medium concentration (0.03%); (c) TPGS as emulsifier with low concentration (0.015%); (d) TPGS asmatrix blended with PLGA with no other emulsifier (ratio for PLGA–TPGS: M1 (2:1), M2 (1:1), M3 (1:2)). The horizontal axis is plottedon log-normal scale.

polymer, oil phase and aqueous phase changes tribution of surfactant stabiliser on the surface, wasaccordingly. In addition, the nanoparticles (samples found to significantly influence the redispersion ofM1–M3) prepared with the mixture of TPGS and the freeze-dried nanoparticles [31,32]. The aggre-PLGA had somewhat smaller size and quite narrow gated nanoparticles may be easily and thoroughlypolydispersity. The polydispersity measurement redispersed to give nanoparticles of smaller andseemed to contradict the SEM images in Fig. 2d, in uniform size as displayed in Fig. 4. Further in-which the large particles might result from aggrega- vestigation is being carried out.tion of small sized particles during the freeze dryingprocess. Generally, there was a tendency for small 3 .2.3. Yield and encapsulation efficiencysized nanoparticles to aggregate during the lyophili- From the results listed in Table 1, it can besation process, which might be one disadvantage of observed that the nanoparticle yield was higher forthe freeze drying technique [30]. The particle size those emulsified by vitamin E TPGS at a mediumwas measured by laser light scattering after suspend- concentration (0.03%); the yield was clearly lowing the particle powder in deionised water by when using TPGS either at high concentrationsonication to form a homogeneous dispersion. The (0.06%) or at low concentration (0.015%). Of thesurface property of nanoparticles, such as the dis- various types of PLGA, it seemed that the yield


increases with decrease of the L/G ratio of PLGA trations up to|20% (w/w), beyond which liquidcopolymer. It was obvious that PLGA (50:50) crystalline phases may form. The amphiphilicconsistently gave higher yield regardless of the characteristic of the TPGS molecule leads to itsemulsifier concentrations and other preparation con- self-association in water when concentration exceedsditions. One main reason might be that when the a threshold known as the critical micelle concen-hydrophobic property increases, the material in the tration (CMC), which is|0.02 wt% in water. Aboveo/w emulsion system with bulky water phase was CMC, TPGS begins to form micelles and continuesrelatively easier to aggregate but did not form a to form relatively low viscosity solutions with waterstable emulsion system. Thus, the PLA frequently until a concentration of|20 wt% is obtained. Whengave lower yield in comparison with PLGA. On the TPGS concentration is above this value, higherother hand, however, there was no clear difference in viscosity liquid crystalline phases start to form. Asthe yield when the polymer concentration varied. its concentration increases, the structure of theFurthermore, as regards the various combinations of TPGS/water liquid crystalline phase evolves gradu-TPGS and PLGA as matrix material for samples ally from isotropic globular micellar to isotropicM1–M3, the yield was decreased when the ratio of cylindrical micellar, mixed isotropic cylindricalTPGS was increased. This should result from the micellar and hexagonal, mixed hexagonal and re-high solubility of TPGS in water. When larger versed hexagonal, reversed globular micellar, andamounts of TPGS were added to the o/w emulsion finally to the lamellar phase [33]. In fabrication ofsystem, more would be dissolved in the water. nanoparticles by the single emulsion solvent evapo-

Another aspect worth noticing is the encapsulation ration/extraction technique, the role of the surfactantefficiency of drug entrapped in nanoparticles. This stabiliser is to stabilise the dispersed-phase dropletshas been previously investigated by the authors [25]. and inhibit coalescence. The amphipathic surfactantsThe present study found that most of the experimen- align themselves at the droplet surface so as total parameters influence the drug encapsulation ef- promote stability by lowering the free energy at theficiency in the nanoparticles, drug loading possibly interface between two phases and resisting coalesc-being one of the most significant factors. It could be ence and flocculation of the nanoparticles. However,understood from the present work that, when the at higher concentration, the state of TPGS in thedrug loading was low, the amount of drug entrapped aqueous dispersing phase has changed and it can notin the nanoparticles would be decreased. In contrast, exert a stabilising effect on the formation of emul-when the drug loading ratio was increased to 10%, sion system, droplet separation and stabilisation, asthe drug encapsulation efficiency could reach 100%. well as nanoparticles hardening. In contrast, it was

Considering the particle size, size distribution and evident that when the concentration was too low, itpolydispersity, the nanoparticle yield and the drug did not act as a emulsifier. Therefore, TPGS wouldencapsulation efficiency, it can be concluded that of not be able to perform as a good surfactant at boththe emulsifier concentrations investigated, 0.02– higher and lower concentrations and could not0.03% vitamin E TPGS concentration has the best produce nanoparticles with ideal properties. Thenanoparticle yield, whereas higher or lower TPGS adhesion, flocculation or aggregation of theconcentration resulted in low yield. Moreover, it nanoparticles could easily occur during the fabrica-could be observed that both high and low TPGS tion process. The optimal concentration, confirmed inconcentration could cause aggregation or adhesion of the present research, was 0.02–0.03 wt%, which isnanoparticles during the fabrication process. The very close to the critical micelle concentrationshape of the nanoparticles also became less spherical (CMC). However, when TPGS was blended togetherand the particles were less homogeneous. with PLGA as the matrix material, none of above-

The result found for TPGS is different from that mentioned effects occurred. This is because all thefor other emulsifiers such as PVA [26–28]. This materials were dissolved in the oil phase and therecould be due to the unique physicochemical prop- was no interaction between TPGS and water. Duringerties of vitamin E TPGS. TPGS is miscible with this process for making nanoparticles, no otherwater and forms solutions with water at concen- emulsifier was needed; the material is self-emulsify-


ing. This is another significant property of TPGS in accordance with those found in the literature [33].drug delivery systems, and deserves further research The higher melting point exhibited for the firstto probe its great potential. heating cycle demonstrated a solid TPGS with

relatively high crystallinity. This crystalline state3 .3. DSC analysis requires more thermal energy for melting than the

lower-energy amorphous states that form duringFig. 5 depicts the DSC thermogram analysis on rapid cooling of the sample. Fig. 5c shows thermo-

the phase transition temperatures (T or T ) of the grams of PLGA nanoparticles fabricated at variousg m

pure material and the drug loaded nanoparticles. Fig. concentrations of TPSG as emulsifier. The graphs are5a gives the glass transition temperature of PLGA very similar each to other and quite similar to that of(50:50) at|45 8C in the second heating cycle. With pure PLGA, which indicates that the influence ofreference to Fig. 5b, the pure vitamin E TPGS shows TPGS as emulsifier was not significant. From Fig. 5dthe endothermic peak of melting point at|41 8C for which shows the thermogram characteristic of thethe first heating cycle and at a lower value of|38 8C nanoparticles produced from the mixture of TPGSfor the second heating cycle. These results were in and PLGA, it can be seen that the range of phase

Fig. 5. DSC thermogram of second heating cycle of fabricated nanoparticles and related material: (a) pure PLGA (50:50); (b) pure TPGS;(c) PLGA nanoparticles with TPGS as emulsifier; (d) nanoparticles with blended mixture of PLGA and TPGS as matrix.


transition temperature appears broader and the data TPGS as emulsifier all had almost similar relativedrop between glass transition of pure PLGA and the ratios amongst the suggested chemical bonds and themelting point of pure TPGS. The shape of the curves ratios were in accordance with those found for pureis very different from that of the melting peak. This vitamin E TPGS. Hence, it can be inferred that themay be due to the joint domination of both PLGA TPGS was mainly distributed over the surface of theand TPGS of the thermal property of the nanoparti- nanoparticles, which meant that the TPGS does actcles. The two materials were blended together and as emulsifier to form and stabilise the nanoparticlesmade each other much more amorphous. Overall, it in the fabrication process. The result showed that thecould be seen that the fabricated nanoparticles all emulsifier could not be fully washed away when thedisplayed phase transition that corresponded to the formed nanoparticles were just washed once or twiceamorphous solid material and did not show any during fabrication [25]. When using TPGS blendedmelting peaks. with PLGA as the matrix material, it was notable

that as more TPGS was used, the proportion of3 .4. Surface analysis C–C/C–H bonds increased while the proportion of

O–C=O bonds decreased. Making a comparison withPolymeric nanoparticles might be taken up by pure TPGS, the result may be attributed to the high

macrophages of the mononuclear phagocytes system, proportion of C–C/C–H bonds in the structure ofwhich may provide an opportunity to deliver drugs to TPGS. This also reflected that in the process ofcertain cellular or tissue sites and may involve forming nanoparticles, there were more TPGS mole-various complicated and simultaneous mechanisms cules distributed on the nanoparticle surfaces. In[34,35]. Among them, interaction between nanoparti- essence, it can be seen that TPGS can act not only ascle and cell play an important role. Surface property emulsifier but also as a novel kind of matrix materialanalysis of drug-loaded nanoparticles should help which has self-emulsifying effects. Due to the am-elucidate the mechanism of particle–cell interactions. phiphilic nature of TPGS, repeated washing wouldThe XPS (ESCA) technique was adopted in the be able to remove the TPGS from the nanoparticlepresent study to investigate the surface chemistry of surfaces.the nanoparticles. The results infer the type ofchemical bonds at the surface of the nanoparticles 3 .5. In vitro releaseand the fabricated nanoparticles and are shown inTable 2 and Fig. 6. Clearly, no matter which type of The in vitro release behaviour of the variousPLGA was used, the nanoparticles fabricated using paclitaxel-loaded nanoparticles is summarised in the

Table 2XPS (C1s) analysis of prepared nanoparticles and involved materials

Sample XPS C1s envelope ratios (%)

C–C/C–H C–O O–C=O

PLGA (50:50) 49.5 26.9 23.6PLGA (75:25) 62.4 20.4 17.2PLA 50.7 26.7 22.5TPGS 57.7 30.0 12.3PLGA (50:50) nanoparticleswith TPGS as emulsifier 54.5 28.0 17.5PLGA (75:25) nanoparticleswith TPGS as emulsifier 58.7 25.4 16.0PLA nanoparticleswith TPGS as emulsifier 58.1 24.4 17.5PLGA-TPGS (2:1) nanoparticles 62.6 23.9 13.5PLGA-TPGS (1:1) nanoparticles 74.4 10.2 15.4PLGA-TPGS (1:2) nanoparticles 78.1 7.9 14.0


Fig. 6. Surface chemistry analysis of C1s with X-ray photoelectron spectroscopy: (a) PLGA (50:50); (b) vitamin E TPGS; (c) PLGA (50:50)nanoparticles using TPGS as emulsifier; and (d) PLGA-TPGS (1:1) nanoparticles.

cumulative percentage release shown in Fig. 7. would be the slowest. Fig. 7b shows the releaseRelease over 1 month was measured. Fig. 7a shows curves when TPGS was used as fabrication materialthe release profiles for nanoparticles fabricated from blended with PLGA; as the ratio of TPGS increased,various types of PLGA using 0.03 wt% TPGS as the relative release rate also increased. This isemulsifier. The initial release burst was prominent because as more TPGS was used, the nanoparticlesfor all three types of polymers during the first day of became more hydrophilic. When exposed to anrelease, being greater than 15%. The release gradual- aqueous environment, the nanoparticles degradedly decreased and remained constant even after 1 more easily. The initial release burst, which was lessmonth. PLGA (75:25) and PLGA (50:50) had simi- than 10%, lasted for about a day, and the druglar release properties, with slightly slower release as release rate then tapered gradually to slower rates. Atime progressed. PLA gave the slowest rate and further point is that the paclitaxel release from the Mextent. This was as expected because of its highly series with TPGS mixed with PLGA as matrix washydrophobic nature and molecular weight (MW), less than that of E series with polymer only aswhich prevented the drug from diffusing from the matrix. One reason might be related to the physico-polymer matrix into the aqueous solution. Size was chemical natures of the material, drug, and surfactantalso a factor. Being the largest of the three batches of as well as their interactions within the nanoparticles.nanoparticles (Table 1), the rate and extent of release Both polymer material and the paclitaxel were quite


Fig. 7. In vitro release curves of paclitaxel loaded nanoparticles prepared under various experiment parameters (a) E5: PLA, E7: PLGA(75:25), E8: PLGA (50:50); (b) Ratio for PLGA-TPGS: M1 (2:1), M2 (1:1), M3 (1:2); (c) PLGA (75:25) concentration – E7: 0.125,E13: 0.188, E14: 0.25; (d) PLGA (50:50) concentration – E8: 0.125, E15: 0.188, E16: 0.25.

hydrophobic and could not dissolve in water. TPGS ones. In accordance with the observations made inwas amphiphilic but its hydrophobic ability was the SEM and AFM studies, the initial burst could begreater than its hydrophilic capability [33]. Therefore due to the diffusion release of paclitaxel distributedwhen TPGS was used as matrix material blended near the surface and in the outer portion of thewith PLGA and no other surfactant substance was nanoparticles. Afterwards, the release rates slowed asadded, the interaction or affinity between the poly- it would require time for the matrix material to erodemer matrix and the drug might be enhanced and thus in the aqueous environment. Thus the release mecha-caused the slower drug release. The DSC study nisms of diffusion, matrix swelling and polymershowed that TPGS and PLGA made each other much erosion, might be the main causes of the releasemore amorphous when the two materials were behaviour [36].blended together, which might also indicate that theinteraction between them greatly increased.

Additionally, the differences in release rates for 4 . Conclusionsthe particles prepared with various polymer con-centrations displayed in Fig. 7c,d were as described The present research proposed a novel formulationbefore for size. The release of drug from smaller by applying vitamin E TPGS either as emulsifier orparticles was clearly faster than that from larger as a component of the matrix material to fabricate


nanoparticles by freeze-dry solvent extraction/ pore. The authors are grateful to final year studentsevaporation technique for controlled release of the Koo Yew Fye and Kiang Tai Wei in the Departmentantineoplastic drug paclitaxel. Investigation of the of Chemical and Environmental Engineering, NUSpreparation, characterisation and in vitro release of for their assistance with experiments.the nanoparticles was carried out. The differentformulations with various ratios of oil phase, aque-ous phase, polymer material and surfactant wereevaluated and optimised. Our results demonstratedR eferencesthat vitamin E TPGS could be an efficient emulsifierfor fabrication of polymeric nanoparticles by the

[1] M.K. Yeh, S.S. Davis, A.G.A. Coombes, Improving proteinsingle emulsion technique, which can achieve excel- delivery from microparticles using blends of poly (D,Llent effects in drug encapsulation efficiency, size and lactide-co-glycolide) and poly (ethylene oxide)-poly (pro-

pylene oxide) copolymers, Pharm. Res. 13 (1996) 1693–size distribution, morphological and physicochemical1698.properties, and in vitro release kinetics of the

[2] R. Jalil, J.R. Nixon, Biodegradable poly (lactic acid) andnanoparticles, and may have the potential to improvepoly (lactide-co-glycolide) microcapsules: problems associ-

nanoparticle adhesion to cells and the hemodynamic ated with preparative techniques and release properties, J.properties of the nanoparticles in the blood flow. Microencapsul. 7 (1990) 297–325.

[3] L. Brannon-Peppas, Recent advances in the use of bio-This is the first time in the literature that TPGSdegradable microparticles and nanoparticles in controlledblended with other biodegradable polymers has beendrug delivery, Int. J. Pharm. 116 (1995) 1–9.used as matrix material for nanoparticle fabrication,

[4] P. Maincent, R. Le Verge, P. Sado, P. Couvreur, J.P. De-and has a self-emulsifying effect. In this research, a vissaguet, Deposition kinetics and oral bioavailability ofdrug encapsulation efficiency as high as 100% has vincamin-loaded poly-alkyl cyanoacrylate nanoparticles, J.been achieved. AFM and SEM revealed the Pharm. Sci. 75 (1986) 955–958.

[5] T.B. Zeltner, T.D. Sweeney, W.A. Skornik, H.A. Feldman,nanoparticles were fine spherical shapes. The particleJ.D. Brain, Retention and clearance of 0.9-mm particlessize and size distribution strongly depended on theinhaled by hamsters during rest or exercise, J. Appl. Physiol.

amount of TPGS added in the fabrication. DSC study 70 (1991) 1137–1145.concluded that when TPGS and PLGA were blended [6] M.J. Cappel, J. Kreuter, Effect of nanoparticles on transder-together to make the nanoparticles, they made each mal drug delivery, J. Microencapsul. 8 (1991) 369–374.

[7] P. Couvreur, P. Tulkens, M. Roland, A. Trouet, P. Speiser,other much more amorphous. The fabricatedNanocapsules: a new type of lysosomotropic carrier, FEBSnanoparticles all displayed phase transition behaviourLett. 84 (2) (1977) 323–326.that corresponded to that of the amorphous solid

[8] A. Astier, B. Doat, M.-J. Ferrer, G. Benoit, J. Fleury, A.material and did not show any melting peaks. XPS Rolland, R. Leverge, Enhancement of adriamycin antitumorinvestigation indicated that the surface of the fabri- activity by its binding with an intracellular sustained-release

form, polymethacrylate nanospheres, in U-937 cells, Cancercated nanoparticles was dominated by the TPGSRes. 48 (1988) 1835–1841.molecules. We can conclude then that vitamin E

[9] V. Labhasetwar, C. Song, R.J. Levy, Nanoparticle drugTPGS has advantages either as emulsifier or asdelivery system for restenosis, Adv. Drug Deliv. Rev. 24

matrix material blended with PLGA for the manufac- (1997) 63–85.ture of nanoparticles for controlled release of pa- [10] P. Beck, J. Kreuter, R. Reszka, I. Fichtner, Influence of

polybutylcyanoacrylate nanoparticles and liposomes on theclitaxel. The formulation can also be applied to makeefficacy and toxicity of the anticancer drug mitoxantrone innanoparticles for clinical administration of othermurine tumour models, J. Microencapsul. 10 (1993) 101–drugs and gene therapy.114.

[11] M. Simeonova, M. Ilarionova, T. Ivanova, C. Konstantinov,D. Todorov, Nanoparticles as drug carriers for vinblastine.Acute toxicity of vinblastine in a free form and associated topolybutylcyanoacrylate nanoparticles, Acta Physiol. Phar-A cknowledgementsmacol. Bulg. 17 (1991) 43–49.

[12] S. Bennis, C. Chapey, P. Couvreur, J. Robert, EnhancedThis research is supported by NUS Grants R-279- cytotoxicity of doxorubicin encapsulated in polyisohex-

000-052-112 and R-279-000-021-112 (PI: A/P Feng ylcyanoacrylate nanospheres against multidrug-resistantSS), National University of Singapore (NUS), Singa- tumour cells in culture, Eur. J. Cancer 30A (1994) 89–93.


[13] C. Verdun, F. Brasseur, H. Vranokx, P. Couvreur, M. Roland, polymeric nanospheres for controlled release of paclitaxelTissue distribution of doxorubicin associated with poly- (Taxol ), J. Controlled Release 80 (1–3) (2002) 129–144.

isohexylcyanoacrylate nanoparticles, Cancer Chemother. [26] H. Jeffery, S.S. Davis, D.T. Ohagan, The preparation andPharmacol. 26 (1990) 13–18. characterization of poly(lactide-co-glycolide) microparticles.

[14] R.A. Jain, The manufacturing techniques of various drug 1. Oil-in-water emulsion solvent evaporation, Int. J. Pharm.loaded biodegradable poly(lactide-co-glycolide) (PLGA) de- 77 (2–3) (1991) 169–175.vices, Biomaterials 21 (2000) 2475–2490. [27] P.D. Scholes, A.G.A. Coombes, L. Illum, S.S. Davis, M.Vert,

[15] P. Johansen, Y. Men, H.P. Merkle, B. Gander, Revisiting M.C. Davies, The preparation of sub-200 nm poly(lactide-PLA/PLGA microspheres: an analysis of their potential in co-glycolyde) microspheres for site-specific drug delivery, J.parenteral vaccination, Eur. J. Pharm. Biopharm. 50 (2000) Controlled Release 25 (1993) 145–153.129–146. [28] L.R. Beck, D.R. Cowsar, D.H. Lewis, R.J. Cosgrove, C.T.

[16] X.S. Wu, Preparation, characterization, and drug delivery Riddle, S.L. Lowry, T. Epperly, A new long-acting injectableapplications of microspheres based on biodegradable lactic / microcapsule system for the administration of progesterone,glycolic acid polymers, in: D.L. Wise et al. (Ed.), En- Fertil. Steril. 31 (5) (1979) 545–551.cyclopedic Handbook of Biomaterials and Bioengineering, [29] Y.M. Wang, H. Sato, I. Adachi, I. Horikoshi, Preparation andMarcel Dekker, New York, 1995, pp. 1151–1200. characterization of poly (lactic-co-glycolic acid) micro-

[17] R. Arshady, Preparation of biodegradable microspheres and spheres for targeted delivery of a novel anticancer agent,microcapsules: 2. Polylactides and related polyesters, J. Taxol, Chem. Pharm. Bull. 44 (1996) 1935–1940.Controlled Release 17 (1991) 1–22. [30] Y.N. Konan, R. Gurny, E. Allemann, Preparation and

[18] A.K. Singla, G. Alka, A. Deepika, Paclitaxel and its formula- characterization of sterile and freeze-dried sub-200 nmtions, Int. J. Pharm. 235 (1-2) (2002) 179–192. nanoparticles, Int. J. Pharm. 233 (2002) 239–252.

[19] E. Tatou, C. Mossiat, V. Maupoil, F. Gabrielle, M. David, L. [31] H. Murakami, Y. Kawashima, T. Niwa, T. Hino, H.Rochette, Effects of cyclosporin and cremophor on working Takeuchi, M. Kobayashi, Influence of the degrees of hydro-rat heart and incidence of myocardial lipid peroxidation, lyzation and polymerization of poly(vinylalcohol) on thePharmacology 52 (1996) 1–7. preparation and properties of poly(DL-lactide-co-glycolide)

[20] R.T. Dorr, Pharmacology and toxicology of Cremophor EL nanoparticle, Int. J. Pharm. 149 (1997) 43–49.diluent, Ann. Pharmacother. 28 (1994) S11–S14. [32] H. Murakami, M. Kobayashi, H. Takeuchi, Y. Kawashima,

[21] M.L. Fjallskog, L. Frii, J. Bergh, Is cremophor, solvent for Preparation of poly (DL-lactide-co-glycolide) nanoparticlespaclitaxel, cytotoxic?, Lancet 342 (1993) 876. by modified spontaneous emulsification solvent diffusion

[22] L. Webster, M. Linsenmeyer, M. Millward, C. Morton, J. method, Int. J. Pharm. 187 (1999) 143–152.Bishop, D. Woodcock, Measurement of cremophor EL [33] Vitamin E TPGS, Properties and Applications, Eastman,following taxol: plasma levels sufficient to reverse drug USA, Publication EFC-226A, October 1998.exclusion mediated by the multidrug-resistant phenotype, J. [34] C. Curvier, L. Roblot-Treupel, J.M. Millot, G. Lizard, S.Natl. Cancer Inst. 85 (1993) 1685–1690. Chevillard, M. Manfait, P. Couvreur, M.F. Poupon, Doxoru-

[23] P. Mankad, J. Spatenka, Z. Slavik, G. Oneil, A. Chester, M. bicin-loaded nanospheres bypass tumour cell multidrugYacoub, Acute effects of cyclosporine and cremophor EL on resistance, Biochem. Pharmacol. 44 (1992) 509–517.endothelial function and vascular smooth-muscle in the [35] C. Verdun, F. Brasseur, H. Vranckx, P. Couvreur, M. Roland,isolated rat-heart, Cardiovasc. Drug Ther. 6 (1992) 77–83. Tissue distribution of doxorubicin associated with poly-

[24] M. Kongshaug, L.S. Cheng, J. Moan, C. Rimington, Inter- isohexylcyanoacrylate nanoparticles, Cancer Chemother.action of Cremophor-EL with human plasma, Int. J. Bio- Pharmacol. 26 (1990) 13–18.chem. 23 (1991) 473–478. [36] S.J. Holland, B.J. Tighe, Biodegradable Polymers, in: D.

[25] L. Mu, S.S. Feng, Vitamin E TPGS used as emulsifier in the Garderton, T. Jones (Eds.), Adv. Pharmaceutical Science,solvent evaporation/extraction technique for fabrication of Vol. 6, Academic Press, New York, 1992, pp. 101–164.

Documents

A novel controlled release formulation for the anticancer drug …drugsanddisease.synthasite.com/resources/novel-controlled... · 2010-07-23 · A novel controlled release formulation